Optical Materials 34 (2012) 2041–2044
Contents lists available at SciVerse ScienceDirect
Optical Materials journal homepage: www.elsevier.com/locate/optmat
Upconversion luminescence properties of nanocrystallite MgAl2O4 spinel doped with Ho3+ and Yb3+ ions A. Watras, P.J. Deren´ ⇑, R. Pa˛zik, K. Maleszka-Bagin´ska Institute for Low Temperature and Structure Research of Polish Academy of Sciences, Okólna 2, 50-422 Wrocław, Poland
a r t i c l e
i n f o
Article history: Available online 6 April 2012 Keywords: Nanocrystallite MgAl2O4 Spinel Upconversion Ho3+ Yb3+
a b s t r a c t The upconversion luminescence spectra of nanocrystallite MgAl2O4 doped with 1% of Ho3+ and 5% of Yb3+ ions after excitation at 980 nm were measured. Inﬂuence of excitation regime either continuous or pulse on upconversion mechanisms was shown. For continuous wave (CW) laser excitation upconversion process is due to phonon assisted Excited State Absorption (ESA). For pulse laser excitation upconversion emission is due to Energy Transfer Upconversion (ETU). Ó 2012 Elsevier B.V. All rights reserved.
1. Introduction There are many papers on upconversion emission in materials doped with Ho3+ and Yb3+ ions [1–5]. One of potential application for such phosphors is using them for enhancing efﬁciency of GaAs photovoltaic cells. These photovoltaic cells are effective for red and green light, but equipped with such phosphor will convert into electricity also the IR part of Solar radiation . MgAl2O4 spinel was chosen as a host for holmium and ytterbium ions. Their physical and mechanical properties are well known. Spinel has high melting point (2135 °C), good thermal shock resistance and mechanical strength . The synthetic nanocrystallite MgAl2O4 was already obtained by the Pechini method . In this work we present for the ﬁrst time (to our knowledge) preliminary studies of upconversion luminescence of nanocrystallite MgAl2O4 spinel doped with Ho3+ and Yb3+ ions. The upconversion phenomenon was investigated for continuous wave (CW) and pulse laser excitation.
2. Experimental Sample was prepared by modiﬁed Pechini method. As the starting materials Mg(CH3COOH)24H2O, AlCl36H2O, Ho2O3 (Alfa Aesar 99.99%) and Yb2O3 (Alfa Aesar 99.99%) were chosen. Citric acid and ⇑ Corresponding author. E-mail address: [email protected]
(P.J. Deren´). 0925-3467/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.optmat.2012.03.016
ethylene glycol were added as a chelating and polymerization agents, respectively. Stoichiometric amounts of lanthanides oxides were dissolved in dilute HNO3 to transfer them into soluble in water nitrate salts and the excess of acid was evaporated at 90 °C. Afterwards all substrates were dissolved in deionized water and mixed with the solution of ethylene glycol and citric acid. The ﬁxed molar ratio of citric acid and ethylene glycol to total chelated metal cations was 5:1. After that ﬁnal homogeneous solution was placed into plastic containers and kept into dryer at 80 °C until brown resin was obtained. Subsequently the resin was sintered at 1100 °C for 8 h in air atmosphere with 10 °C/min temperature rise step. Structure studies of the obtained nanocrystallite powder were characterized using XRD powder diffraction. The XRD were measured on Panalitycal X’pert equipped with Cu Ka lamp. For CW excitation and power dependence measurements we have used ﬁber coupled 975 nm laser diode with 4 W output power. The end of the ﬁber was at 0.5 cm distance from sample, no lens was used and estimated spot diameter was 2.5 mm, accordingly when laser diode was set at 4 W (at its maximum) the power density was 82 W/cm2. For pulse excitation we have used 980 nm line from tunable Ti:Sapphire laser pumped by second harmonic Nd:YAG. Laser light was focused to a 2 mm diameter spot at the sample by a 4 cm focal length lens. Laser pulse energy was set to 2 mJ and pulse duration was 10 ns, the repetition rate was 10 Hz. The emission was analyzed by Jobin Yvon THR1000 monochromator equipped with Hamamatsu R928 photomultiplier and CCD camera. To measure emission decay proﬁles we used digital oscilloscope LeCroy WaveSurfer 400.
A. Watras et al. / Optical Materials 34 (2012) 2041–2044
Fig. 1. XRD spectrum of nanocrystallite MgAl2O4 spinel doped with 1% of Ho3+ and 5% of Yb3+ ions.
Fig. 2. Upconversion emission spectra of MgAl2O4 doped with 1% of Ho3+ and 5% of Yb3+ ions at 300 K.
Fig. 3. Upconversion emission spectra of MgAl2O4 doped with 1% of Ho3+ and 5% of Yb3+ ions at 77 K.
3. Results and discussion The XRD spectrum of MgAl2O4 doped with 1% of Ho3+ ions and 5% of Yb3+ ions is presented in Fig. 1. It can be seen that sample has
structure without any additional phases. The average grain size estimated with Williamson–Hall method is 11 nm. The upconversion luminescence spectrum measured at 300 K is presented in Fig. 2. The sample was excited at 980 nm by CW and pulsed laser. Solid line shows in Fig. 2 emission obtained after CW excitation and dashed after pulse excitation. One can see that for CW excitation there are only two emission bands; one green centered at 550 nm assigned to the (5F4, 5S2) ? 5I8 transition and second red centered at 660 nm and assigned to the 5F5 ? 5I8 transition. For pulse excitation there is additional band with maximum at 585 nm, which corresponds to the (3K7, 5G4) ? 5I6 transition. One can expect to observe transitions from the 5G4 level to the ground and ﬁrst excited level, the 5I8 and the 5I7, respectively. But we did not observe them, because they are very weak. Taking into account Squared Reduced Matrix Elements and Omega parameters obtained for other oxides doped with Ho(III) , it was possible to calculate the so called transition strengths of above mentioned transitions. After that we can formulate a rule of thumb, that the 5 G4 ? 5I6 transition is much stronger than the 5G4 ? 5I8 and the 5 G4 ? 5I7 one. By the way the latter was covered by second harmonic of the excitation line, therefore not seen. We have to stress that spectra of the sample excited by pulse laser differ signiﬁcantly from those excited by CW radiation, although in both cases the samples were excited at the same wavelength. The low temperature make this difference even more striking (see Fig. 3), the small band at 585 nm disappears and the red band become almost invisible for pulse excitation. It is evident from the experimental data, that the mechanisms of upconversion emission for CW and pulse excitation are different. The processes that occur with CW excitation will be discussed ﬁrst. Excitation wavelength match well the absorption of Yb3+ ions therefore due to high absorption cross-section of Yb3+ in spinel incident photons are well absorbed. Then the energy is transferred to the 5I5 and 5I6 levels of Ho3+ ions (see Fig. 4). The 5I7 level is populated by nonradiative transitions from the 5I6 one. The emitting (5F4, 5S2) and 5F5 levels are populated by excited state absorption (ESA) from the 5I6 and 5I7 levels, respectively. At 300 K the red emission is approximately two times more intense than green one, but at 77 K green emission has the same intensity as the red one (see Fig. 3). This can be explained by very efﬁcient at 300 K multi-phonon energy transfer from the 5I6 and the 5S2 level. Transfer from the 5I6 level populates the 5I7 one and thus the 5 I7 ? 5F5 ESA becomes more effective. Second mentioned above transfer feeds directly the 5F5 level. At 77 K multiphonon transfer is much less efﬁcient and therefore the green emission becomes relatively stronger. For pulse excitation the upconversion mechanism is different, the emission is observed due to so called energy transfer upconversion (ETU). Pump energy excites mainly the green emission (see Fig. 2). Incident photons are also well absorbed by the Yb3+ ions then energy is transferred directly to the 5I5, 5F4 and 3H6 + 3D2 levels, which is indicated in Fig. 5 by WYH1, WYH2 and WYH3 arrows, respectively. The emitting levels are populated via multiphonon nonradiative transitions. The 5F5 level is populated with assistance of cross relaxation process: (5S2, 5I8) ? (5I4, 5I7) (dotted arrows in Fig. 5.), which populates 5I7 and makes the ESA process possible. The cross relaxation as well as last energy transfer step: WYH3 must be phonon assisted because at 77 K emission from the 5G4 quintet is not observed and the red emission from 5F5 level is barely seen. Dependence of intensity of upconversion emission on excitation power (CW regime) in MgAl2O4:(Ho3+, Yb3+) is plotted in log–log scale in Fig. 6. For the green emission two different slopes can be seen for low and high pump power. For low pump power (from
A. Watras et al. / Optical Materials 34 (2012) 2041–2044
Fig. 4. Ho3+ and Yb3+ energy levels and excitation paths of the upconverted emission in MgAl2O4 for CW excitation.
Fig. 5. Ho3+ and Yb3+ energy levels and excitation paths of the upconverted emission in MgAl2O4 for pulse excitation.
71 to 1180 mW) the slope is 1.3 and for high pump power (above 1180 mW) the slope is 0.8. For red emission there is only one slope equal 1.1. These values for both the green and the red anti-Stoke’s emission are far from expected 2, but according to Pollnau et al.  intensity of upconversion emission has dependence of absorbed power close to P1 when rate of upconversion (both ETU and ESA) is strong and emission decays predominantly to ground state. For the green emission above 1180 mW saturation is visible. The decay curve of green upconversion emission measured at 300 K is presented in Fig. 7. In the inset the rise of the upconversion emission after excitation pulse is shown, the rise time is equal
320 ns. Decay curve is not single exponential, two decay times were extracted from the curve; s1 = 3 ls and s2 = 9 ls. Decay times are unusually short for Ho3+. Such short values are due to inﬂuence of host i.e. defects caused by mismatch in valence between Mg2+ and dopants or small diameter of the nanocrystallite. In such small nanocrystallite most of the doped ions enter into unit cells located on the nanocrystallite’s surface. Dopants see nonequivalent crystal ﬁeld, which is also well demonstrated by the multi exponential character of decays proﬁles. As it is observed in other nanomaterials there are many sites of doped ions. Rise time conﬁrm ETU mechanism in the pulse excitation because for the ESA the upconversion emission would appear together with the excitation pulse.
A. Watras et al. / Optical Materials 34 (2012) 2041–2044
4. Conclusions Nanocrystallite with 11 nm diameter of MgAl2O4 doped with Ho3+ and Yb3+ ions were synthesized. The upconversion luminescence spectra were observed at 77 and 300 K. Decay times were measured at room temperature. It was shown that upconversion spectra strongly depend on excitation regime. For CW excitation anti-Stokes emission is observed due to ESA and the red emission dominates the spectrum. For pulse excitation green anti-Stokes emission is observed as a result of ETU. Both mechanisms are very effective. Decay proﬁles are multi exponential and decay time of green upconverted emission is very short for Ho3+ ions. This is due to multi site character of the host and very small diameter of MgAl2O4 nanocrystallite and high concentration of dopants. Acknowledgments This work was ﬁnancially supported by the Ministry of Science and Higher Education under Grant No. N N507 372335, which is greatly acknowledged. The authors wish to thank to Dr. V. Kinzhybalo for XRD measurements. References     Fig. 6. The plot of the green (upper) and red (lower) upconversion emission in function of CW pump power.
     
Fig. 7. Decay curve of green upconversion emission. In the inset rise of the emission after excitation pulse is shown.
X. Huang, G. Wang, J. Lumin. 130 (2010) 1702–1707. X.P. Chen, W.J. Zhang, Q.Y. Zhang, Physica B 406 (2011) 1248–1252. E. Osiac, I. Sokólska, S. Kück, J. Lumin. 94–95 (2001) 289–292. J.C. Boyer, F. Vetrone, J.A. Capobianco, A. Speghini, M. Bettinelli, Chem. Phys. Lett. 390 (2004) 403–407. M. Liu, S. Wang, D. Tang, L. Chen, J. Ma, J. Rare Earths 27 (2009) 66–70. X.B. Giang, W.H. Du, X.L. Chang, H.R. Yuan, Sol. Energy Mater. Sol. Cells 68 (2001) 97–103. M.J. Iqbal, S. Farooq, Mater. Sci. Eng. B 136 (2007) 140–147. P. Głuchowski, R. Pa˛zik, D. Hreniak, W. Stre˛k, Chem. Phys. 358 (2009) 52–56. A.A. Kaminskii, Crystalline Lasers: Physical Processes and Operating Schemes, CRC Press, Boca Raton, 1996. M. Pollnau, D.R. Gamelin, S.R. Lüthi, H.U. Güdel, M.P. Hehlen, Phys. Rev. B 61 (2000) 3337–3346.